 |
 |
Adrian, (1) Those contrasts compare one task against the average of the other three, not necessarily that interesting. You can do zeros(1,11) 1 -1 0 0 to compare C1 versus C2, other comparison would follow the same idea. See #2 for how this is derived. (2) Contrasts do not need to sum to 0 (see below); however, contrasts that do not sum to zero must include both condition and subject effects. If you want to see the logic behind this, here is a more complicated example: This is for a design with 18 subjects in group 1, 9 subjects in group 2, 2 group terms and 2 conditions: Start with the simpliest element, single subject in a single condition, build its contrast, repeat for all subjects and conditions, and then combine the ones you want. S1G1C1=[1 zeros(1,26) 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0] S1G1C2=[1 zeros(1,26) 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0] .... Now average your G1C1 and by summing and dividing by the number of subjects, you'd get G1C1=[ones(1,18)/18 zeros(1,9) 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0] and G1C2=[ones(1,18)/18 zeros(1,9) 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0] and G2C1=[zeros(1,18) ones(1,9)/9 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0] and G2C2=[zeros(1,18) ones(1,9)/9 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0] Now subtract G1C1-G1C2 AND G2C2-G2C1 G1C1-G1C2=[zeros(1,27) 0 0 1 -1 0 0 0 0 0 1 -1 0 0 0 0 0 0 0 0 0 0 0 0] and G2C1-G2C2=[zeros(1,27) 0 0 1 -1 0 0 0 0 0 0 0 0 0 0 0 0 1 -1 0 0 0 0 0] Now subtract these two: Interaction contrast=[zeros(1,27) 0 0 0 0 0 0 0 0 0 1 -1 0 0 0 0 0 -1 1 0 0 0 0 0] And enter it into SPM. Best Regards, Donald McLaren ================= D.G. McLaren, Ph.D. Postdoctoral Research Fellow, GRECC, Bedford VA Research Fellow, Department of Neurology, Massachusetts General Hospital and Harvard Medical School Office: (773) 406-2464 ===================== This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is intended only for the use of the individual or entity named above. If the reader of the e-mail is not the intended recipient or the employee or agent responsible for delivering it to the intended recipient, you are hereby notified that you are in possession of confidential and privileged information. Any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited and may be unlawful. If you have received this e-mail unintentionally, please immediately notify the sender via telephone at (773) 406-2464 or email. On Thu, Nov 25, 2010 at 7:50 AM, Jonathan Peelle <[log in to unmask]> wrote: > Hi Adrian > > > 1. Are the above contrasts correct. > > Yes, they seem fine. > > > > 2. Why do the contrasts have to sum up to 0? I thought the F-contrast > should be [zeros(4, 11) eye(4)], but SPM says "invalid contrast". > > When you have columns for subject effects, it's not possible to > directly estimate the mean for certain conditions. I.e., if you look > at the indicators along the bottom of the design matrix, they will be > gray. This means you can't do the equivalent of a "[1 0...]" contrast > with these columns, and thus an eye(4) contrast is similarly invalid > (because each row contains an invalid contrast). Comparing conditions > is still fine (as you've discovered). > > > > 3. Can I simply multiply my contrasts by -1 to test for "negative" > effects? > > Yes. > > Hope this helps! > > Jonathan >